ScalingIntelligence/swe-bench-verified-codebase-content-staging

该数据集包含两个主要配置：file_content和problem_files。file_content配置包含文件的哈希值和内容，problem_files配置包含实例ID和文件列表，文件列表中包含文件的哈希值和路径。每个配置都有训练集和测试集，分别提供了字节数和示例数。此外，还提供了每个配置的下载大小和数据集大小。

The dataset contains two main configurations: file_content and problem_files. The file_content configuration includes the hash and content of files, while the problem_files configuration includes instance IDs and a list of files, where each file contains a content hash and file path. Each configuration has a training set and a test set, with the number of bytes and examples provided for each. Additionally, the download size and dataset size for each configuration are provided.